Analysis and optimization of an 8 bar mechanism

This paper deals with an 8-bar mechanism, which finds its application in a deep drawing press. It is required in the process of deep drawing that the slider maintains an almost constant velocity during a portion of the return stroke. The approach for the kinematic simulation is based on loop closure analysis, which has been performed to derive expressions for slider displacement, slider velocity and slider acceleration. An extensive sensitivity analysis is carried out to identify key parameters affecting the mechanism. A modified Multiple Linear Regression (MLR) method is introduced, by splitting the displacement plot into three regions to consolidate the results, in order to account for an otherwise erroneous path generated by using the conventional MLR equation treating the plot as a single region. As an alternative method Artificial neural network (ANN) is introduced for consolidation of data. It is shown that ANN also proves to be a better and easier procedure for data consolidation. The consolidated data is used to derive the objective function for optimization. Genetic Algorithm is adopted for optimizing the mechanism with respect to the sensitive link lengths, satisfying the objective function.